Most of the health hazards in fried foods are related to unqualified frying oil and excessive oil content. In this study, the feasibility of using low-field nuclear magnetic resonance techniques (LF-NMR) for analysis of the water and oil contents in French fries, as well as simultaneous evaluation of frying oil quality during deep frying, was investigated. Three proton populations were identified and successfully assigned to water and oil relaxation signals. Significant correlation between the T2 relaxation parameters ( Awater and RCoil) and the water and oil content was acquired. MRI could visualize the changes of signal intensity and spatial distribution, as well as the internal structural changes during frying. Using the correlation model built by multiple regression analysis, the total polar compounds content of the frying oil could be successfully predicted by LF-NMR relaxation characteristics, which indicates that LF-NMR was an effective method to monitor the quality of frying oil.
Keywords: French fries; LF-NMR; MRI; TPC; quantitative analysis.